Google News API, Google News Search API and RSS Alternative for Real-Time JSON
Send a keyword or topic and the US edition, and get the live Google News page back as JSON: every story's headline, publisher, publish time, link and the cluster it sits in. Schedule your watchlist to run every 15 minutes and land in your webhook. One results page is one request, at a flat rate.
- Status
- ...
- Elapsed
- ... ms
- Exit country
- ...
- Browser used
- ...
This console sends a real request and shows you the real response. Five live fetches per session, and the sample targets run as often as you like.
There is no official Google News API. Here is what works instead
Google shut down its News Search API more than a decade ago and never replaced it. Today a Google News API means one of three things: the unsupported RSS feeds, a news aggregator that indexes publishers itself, or an API that fetches the real Google News pages and parses them. Only the last one gives you what Google News actually ranks right now.
That distinction matters to the people who pay for this. A PR agency reporting share of voice to a client wants the stories Google puts in front of readers, in Google's order, not a keyword match across 80,000 sources. A trading desk watching for a ticker wants the moment a story enters the top cluster. A brand team wants to know which outlet broke the story that everyone else rewrote.
WebScrap fetches Google News search, topic and headline pages from a US location with the edition you set, keeps the requests unblocked, and returns the stories as typed JSON. You keep the data and the alerts. We keep the fetching working.
Google News API options compared
Each of these is the right answer for somebody. The deciding questions are whether you need Google's ranking or just news coverage, how far back you need to search, and whether the data ends up in something a customer pays for.
| Option | Good at | Where it falls short |
|---|---|---|
| Google News RSS feeds | No cost, no key, fine for a prototype or a personal reader | Capped list with no pagination, redirect links instead of publisher URLs, no service agreement, and rate limits once you poll hard from one server |
| News aggregator APIs (News API, GNews) | Full text search across tens of thousands of sources, historical archives back several years, clean article fields | Their own index and ranking, not Google's, so the top stories you get are not the ones Google readers see |
| SERP API specialists | Parsed Google News results with little setup, many other Google verticals on the same key | Priced per search on tiered bundles, from about 9 to 25 USD per 1,000 on the entry plans of the best known vendor |
| In-house scraper | Full control of what gets fetched and parsed | You own the proxies, the blocks, the edition parameters and a parser that breaks when the layout changes |
| WebScrap Google News API | Live Google News pages as JSON from a US location, the same flat rate as any other page, scheduled runs to your webhook | No archive. You get what Google News shows now and store the history yourself. For years of back issues, an aggregator is the better buy |
What 1,000 Google News searches cost, provider by provider
Prices read from each provider's public pricing page in September 2026, converted to the one unit that matters for monitoring: dollars per 1,000 searches or requests. Aggregators are included because buyers compare them, even though they answer a different question.
| Provider and plan | Per month | Included | Per 1,000 |
|---|---|---|---|
| SerpApi Developer | 75 USD | 5,000 searches | 15.00 USD |
| SerpApi Big Data | 275 USD | 30,000 searches | about 9.17 USD |
| HasData Startup, annual billing | 49 USD | 20,000 news searches | about 2.45 USD |
| HasData Growth, annual billing | 208 USD | 300,000 news searches | about 0.69 USD |
| News API Business (aggregator) | 449 USD | 250,000 requests | about 1.80 USD |
| WebScrap Starter | 49 USD | 50,000 requests | 0.98 USD |
| WebScrap Growth | 149 USD | 250,000 requests | 0.60 USD |
| WebScrap Scale | 499 USD | 1,500,000 requests | 0.33 USD |
Honest reading: HasData's annual plans sit in the same range as ours for Google News alone, and if a single news feed is the only thing you will ever fetch, price them too. The difference shows up the moment the job grows. Monitoring rarely stops at the headline. You resolve the link, fetch the article, check the Google search results for the same story, and on WebScrap every one of those is the same request at the same rate. News API's free Developer plan, for the record, is licensed for development only and has a 24 hour delay, so it is not an option for a product.
What comes back for every story
You ask for fields in plain English with the describe parameter, or pass a schema, and each story comes back typed. The default set covers what monitoring and alerting actually use.
- Headline, exactly as Google News shows it
- Publisher name, so you can weight tier one outlets separately from syndication
- Publish time as shown, normalized to an ISO timestamp with the fetch time next to it
- The Google News link, plus the resolved publisher URL when you ask for it
- Position on the page and the story cluster it belongs to, so you can tell the lead story from the rewrites
- Thumbnail URL when the page shows one
How a Google News watchlist runs in four steps
01
List the queries
Brand names, executives, tickers, competitors, product names. Each one is a Google News search URL with the US English edition set.
02
Choose the fields
Send the URL to /v1/scrape with the fields you want described in plain English. The response carries them typed, ready for your database.
03
Schedule it
On Growth and above, save the set as a job that runs every 15 minutes, hourly or daily and posts results to your webhook.
04
Alert on what is new
Compare each run with the last by resolved URL and alert only on new stories. Failed fetches are retried and never billed, so gaps are rare.
Who pays for Google News data
PR and communications agencies
Coverage reports and share of voice for every client, built from what Google News readers saw, on one invoice rather than a media monitoring seat per client.
Finance and trading teams
Ticker and company watchlists polled every few minutes, with the publish time and the moment the story entered the top cluster stored side by side.
Media intelligence products
News as a feature of your own SaaS without a proxy fleet. Your cost per monitored keyword is a number you can price your tiers against.
AI and research teams
Fresh headlines and article text to ground an assistant or feed a classifier, with the source and time on every record so answers can cite them.
Three traps that break home-built Google News scrapers
Redirect links. Every link on Google News and in its RSS feeds points to a news.google.com address, not to the publisher. If you store those, you cannot deduplicate a story that two queries both found, and you cannot fetch the article. Resolve each link to the publisher URL once, when the story first appears, and key your records on that.
The wrong edition. Google News ranks differently per country and language edition. A server in Frankfurt fetching without edition parameters gets German results for an English brand name. Set the US English edition explicitly and fetch from a US location, every run, or your share of voice numbers will drift for reasons that have nothing to do with coverage.
Polling from one address. A monitoring job that hits Google every few minutes from the same IP is the textbook pattern that gets throttled. The first sign is not an error but an empty or partial page that your parser happily reads as "no new stories". Rotate the exit address and treat an empty result on a busy keyword as a failed fetch, not as silence. Our rotating proxy pool does this on every request.
Sizing a monitoring budget
Requests per month are queries times checks per day times 30. One results page per check covers the stories most monitoring needs. Here is how common watchlists map to plans.
| Watchlist | Checks per day | Requests per month | Plan it fits |
|---|---|---|---|
| 200 client brand terms, hourly | 24 | 144,000 | Growth, 149 USD |
| 500 terms, hourly | 24 | 360,000 | Scale, 499 USD |
| 100 tickers, every 15 minutes | 96 | 288,000 | Scale, 499 USD |
| 1,000 long tail terms, twice daily | 2 | 60,000 | Growth, 149 USD |
| 150 terms, three times a day | 3 | 13,500 | Starter, 49 USD |
Add one request per article if you fetch full text. A watchlist that surfaces 300 genuinely new stories a day adds about 9,000 requests a month, which rarely moves the plan. Scheduled jobs are on Growth and Scale. Starter covers on demand requests you trigger yourself, which is how most teams run the first month while they tune the query list.
Google News API questions
Is there a Google News API?
No, not an official one. Google retired its News Search API years ago and has not replaced it. Every Google News API you can call today is a third party service that either fetches Google News pages and parses them, or indexes publishers directly and borrows the name. Google still publishes RSS feeds, but they are not a supported API.
How much does a Google News API cost?
On WebScrap one Google News results page is one request, billed at 0.98 USD per 1,000 on Starter, 0.60 on Growth and 0.33 on Scale. Checking 500 keywords every hour is about 360,000 requests a month, which fits the 499 USD Scale plan. SERP API specialists charge roughly 9 to 25 USD per 1,000 news searches on their entry tiers.
Is the Google News API deprecated?
Yes. The original Google News Search API was deprecated in 2011 along with the rest of Google's AJAX Search APIs and later switched off. Code and tutorials that still reference it will not work. The replacements are Google News RSS feeds, news aggregator APIs, and APIs that fetch the live Google News pages.
How do I get a Google News API key?
You cannot get one from Google, because Google does not offer a News API. The key you need comes from whichever third party service you use. On WebScrap you create an account, the key is in your dashboard straight away, and the same key covers Google News, Google Search and any other page you fetch.
Can I use the Google News RSS feed as an API?
For a prototype, yes. For production it is fragile. The feed returns a capped list of recent items with no pagination, links point to news.google.com redirect URLs rather than the publisher, there is no service agreement, and heavy polling from one server gets rate limited. Teams usually move off it once the feed feeds something customers pay for.
Can you scrape Google News?
Yes. Google News search and topic pages are public and render headlines, publishers and times in the page. The work is keeping requests unblocked at volume, setting the right edition, and parsing a layout that changes. A scraping API handles the proxies, retries and parsing and returns the stories as JSON.
What is the best Google News API?
It depends on the job. For Google's own ranking of what is top right now, pick an API that fetches live Google News pages. For full text search across publishers with years of history, a news aggregator such as News API or GNews fits better. For very high volume, compare cost per 1,000 searches, which ranges from under 1 USD to over 20 USD.
How do I get the full article text from Google News?
Google News only shows the headline, publisher and a link. Resolve the link to the publisher's URL, then fetch that page and extract the body. On WebScrap both steps use the same endpoint and the same flat rate, so a search plus ten articles is eleven requests. Paywalled articles return only what the publisher shows logged out visitors.
Can I get Google News results for a specific US state or city?
Google News editions are set per country and language, such as the US English edition, and local stories surface through location topics and local queries. Send the request from a US location with the edition parameters set, and add the city or state to the query for local coverage monitoring.
How often can I check Google News for new stories?
As often as your plan's volume allows. Most monitoring teams check priority keywords every 15 to 60 minutes and long tail terms a few times a day. Each check is one request per results page, so 100 keywords every 15 minutes is about 288,000 requests a month, inside the Scale plan.
Put your first news watchlist on a schedule
Send a Google News search and get headlines, publishers and times back as JSON. Flat 0.98 USD per 1,000 requests, falling to 0.33 on Scale, failed fetches never billed.